R1-2509145 discussion

Channel coding for 6GR interface

From MediaTek
Status: noted
WI: FS_6G_Radio
Agenda: 11.4.1
Release: Rel-20
Source: 3gpp.org ↗

Summary

MediaTek presents 10 observations and numerous proposals for 6G channel coding enhancements, covering QC-LDPC data channel improvements targeting 2-4x peak data rate gain over 5G and Polar code enhancements for control channels addressing payload size scalability, early termination, RNTI false alarm reduction, and high-order QAM support.

Position

MediaTek proposes studying 6G QC-LDPC designs capable of achieving 2-4x peak data rate over 5G, emphasizing that simply scaling existing 5G LDPC via additional parallel decoders is area-inefficient and undesirable. They introduce decoding cycle metrics (I_total = I_iter × I_cycle) to evaluate throughput-latency tradeoffs, identifying MCS=20 as the peak data rate bottleneck for BG1. For Polar code enhancements, they propose a new data integrity check mechanism to improve payload size scalability beyond the current 140-bit limit and early termination rate, noting that 5G distributed CRC design provides limited early termination benefit with >50% information bits processed before decoding termination. MediaTek requires studying new RNTI scrambling mechanisms to reduce RNTI-FA rate, presenting a specific proposal to shift CRC bits scrambled by RNTI to the payload beginning, and provides detailed simulation showing 5G design encounters >90% FA rate for certain payload sizes. They propose studying high-order QAM polar codes for both UCI performance enhancement and DCI spectral efficiency, presenting multi-level coding (MLC) with joint shaping results showing >0.5dB gain over 5G BICM. They support pair-wise orthogonality structures and BG size adaptation for different code rates as specific design directions.

Key proposals

Your notes

Private to your account